TRANSACTlOiNS 01!' SilCTlO.N G. 849 



mechanical power by generating it at central stations and distributing it, and, 

 lastly, how far means of transporting energy are likely to make available cheaper 

 sources of energy than steam-power. 



Let us go back for a moment to James Watt. Q'he most distinct feature about 

 the invention of the steam-engine is that it arose out of studies of such ques- 

 tions as the relation of pressure and temperature of steam, the heat absorbed 

 in producing it, and its volume at ditferent pressures. 



Armed with this knowledge. Watt was able to determine that the quantity of 

 steam used in a model atmospheric engine was enormously greater than that "due 

 to the volume described by the piston. There was waste or loss. To discover the 

 loss was to get on the path of hnding a remedy. The separate condenser, by 

 diminishing cylinder condensation, annulled a great part of the loss. So great was 

 Watt's insight into the action of the engine that he was able to leave it so perfect 

 that, except in one respect, little remained for succeeding engine builders, e.xcept to 

 perfect the machines for its manufacture, to improve its details, and to adapt it to 

 new purposes. Now it very early became clear that there were two directions of 

 advance which ought to secure greater economy. Simple mechanical indications 

 showed that increased expansion ought to insure increased economy. Thermodynamic 

 considerations indicated that higher pressures, involving a greater temperature range 

 of working, ought to secure greater economy. But in attempting to advance in either 

 of these directions, engineers were more or less disappointed. Some of Watt's eno-ines 

 worked with 5 lb. of coal per indicated horse-power per hour. Many engines with 

 greater pressures and longer expansions have done but little better. The history of 

 steam-engine improvement for a quarter of a century has been an attempt to secure 

 the advantages of high pressures and high ratios of expansion. The difficulty to be 

 overcome ha.s proved to be due to the same cause as the inefficiency of Watt's 

 model engine. The separate condenser diminished, but it did not annul the action 

 of the cylinder wall. The first experiments which really startled thoughtful steam 

 engineers were those made by 5lr. Isherwood between 1860 and 1865. Mr. 

 Isherwood showed that in engines such as those then in use in the United States 

 Navy, with the large cylinders and low speeds then prevalent, any expansion of 

 the steam beyond three times led, not to an increased economy, but to an increased 

 consumption of steam. Very little later than this M. Hirn undertook, in 1871-5, his 

 classical researches on the action of the steam in an engine of about 160 indicated 

 horse-power. Experiments of greater accuracy or completeness, or of greater insight 

 into the conditions which were important, have never since been made, and Ilirn with 

 Lis assistants, MM. Hallaner and Dwelshauvers Dery, has determined, once for all, 

 the whole method of a perfect steam-engine trial. M. Hirn was the first to clearly 

 realise that the indicator gives the means of determining the steam present in the 

 cylinder during every period of the cycle of the engine. Consequently, superheating 

 in ordinary cases being out of the question, we have the means of determining the 

 heat present and the heat already converted into work. The heat delivered into 

 the engine is known from boiler measurements, combined with calorimetric tests of 

 the quality of the steam, tests which Hirn was the first to undertake. The balance 

 or heat unaccounted for is, tlien, a waste or loss due to causes which have to be 

 investigated. Hirn originated a complete method of analysis of an engine test, 

 sliowing at every stage of the operation the heat accounted for and a balance of 

 heat unaccounted for ; and the latter proved to be a very considerable quantity. 



Meanwhile theoretical writers, especially Rankine and Clausius, had been per- 

 fecting a thermodynamic theory of the steam-engine, based primarily on the 

 remarkable and irrefragable principle of Carnot. Tbe result of Ilirn's analj'sis was 

 to show that these theories, applied to the actual steam-engine, were liable to lead 

 to errors of 50 or 60 per cent., the single false assumption made being that the 

 interaction between the walls of the cylinder and the steam was an action small 

 enough to be negligeable. 



In this country Mr. Mair Ilumley, following Ilirn's method, ma^le a series of 

 experiments on actual engines with great care and accuracy and completeness. All 

 these experiments demonstrated the fact of a large initial condensation of steam on 

 the walls of the cylinder, alike in jacketed and unjacketed engines. This con- 



1892. 3 1 



